Optimization of Transportation Problems with Transshipment by the Hopfield Neural Network

The Hopfield neural network is a recurrent neural network used to associate ideas and recollections and optimize computation.Transportation problems are a special type of linear programming problems.Combing the optimization function of the Hopfield neural network and the characteristics of transshipment problems under actual situations,constraint boundaries are made fuzzy.Continuous Hopfield neural network circuits are designed,aiming at transportation problems with transshipment.The energy function concept of the neural network is made use of to ascertain the parameters of the neural network circuit and to prove the stability of the system.Transportation optimization is turned into seeking the balancing point of the neural network,i.e.,the attracting point.The simulated anneal algorithm is adopted to help the system reach global optimal solutions and avoid system convergence to local sub-optimal solutions.The solving speed is very quick and of good real-time nature.A novel optimization approach to linear and nonlinear programming is put forward.Computer simulations prove the effectiveness and practicability of the Hopfield neural network.